Bull Eng Geol Environ (2010) 69:523–532 DOI 10.1007/s10064-009-0260-8
ORIGINAL PAPER
Risk assessment based selection of rock TBM for adverse geological conditions using Fuzzy-AHP Jafar Khademi Hamidi • Kourosh Shahriar • Bahram Rezai • Jamal Rostami • Hadi Bejari
Received: 28 November 2008 / Accepted: 15 November 2009 / Published online: 8 January 2010 Ó Springer-Verlag 2009
Abstract Uncertainty in geological and geotechnical parameters is inevitable as it is never possible to accurately determine every aspect of the ground conditions along the tunnel alignment during site investigations. As a consequence, there are many potential sources of geotechnical risk in both mechanized and conventional tunneling techniques. Problems such as encountering crushed and fault zones with running gouge, tunnel face and wall instabilities in blocky ground, large quantities of water inflow and converging tunnel sections are the principal geotechnical risks which affect the selection of an appropriate tunnel boring machine. The paper discusses the use of a fuzzy analytical hierarchy process as an efficient means of decision-making, which can take into account the different criteria pertinent to proper machine selection for anticipated ground conditions. Keywords Fuzzy-AHP TBM selection Adverse geological conditions Geotechnical risk TBM performance Zagros long tunnel Re´sume´ L’incertitude sur les parame`tres ge´ologiques et ge´otechniques est ine´vitable car il n’est jamais possible de de´terminer pre´cise´ment chaque aspect des conditions de J. Khademi Hamidi (&) K. Shahriar B. Rezai Department of Mining and Metallurgical Engineering, Amirkabir University of Technology, Hafez 424, P.O. Box 15875-4413, Tehran, Iran e-mail:
[email protected];
[email protected] J. Rostami Department of Energy and Mineral Engineering, Pennsylvania State University, University Park, PA, USA H. Bejari Member of Young Researchers Club, Islamic Azad University of Ghaemshahr, Ghaemshahr, Iran
terrain pendant les reconnaissances le long d’un trace´ de tunnel. En conse´quence, il y a de nombreuses sources potentielles de risque ge´otechnique a` la fois dans la mise en œuvre des techniques de creusement traditionnelles ou me´canise´es. Des proble`mes tels que la rencontre de zones broye´es et faille´es avec des remplissages meubles, des instabilite´s au front et sur les parements dans des terrains fracture´s, des venues d’eau importantes et des phe´nome`nes de convergence sont les principaux risques ge´otechniques qui affectent le choix d’un tunnelier adapte´ au terrain. L’article discute de l’utilisation d’une me´thode base´e sur l’approche Fuzzy AHP dans la prise de de´cision, qui puisse bien prendre en compte les diffe´rents crite`res pertinents pour le choix d’un tunnelier adapte´ aux conditions de terrain. Mots cle´s Fuzzy AHP Choix detunnelier Conditions ge´ologiques difficiles Risque ge´otechnique Performance de tunnelier Tunnel du Zagros
Introduction Selection of the appropriate tunnel boring machine (TBM) for use in various ground conditions is a very delicate and important part of planning a tunneling project. This is because this decision is almost irreversible and once the machine is placed in the ground, there is almost no way to pull it out and any major modifications and retrofitting is very time consuming and costly, if not impossible. In the meantime, if the machine type is not suitable for the ground conditions, it can cause major delays, could be detrimental to the safety of the crew and personnel, and ultimately could bring the project to a halt. Obviously, accurate characterization of the ground using surface and subsurface investigation will allow the
123
524
designers to foresee the potential problems and select a machine that can cope with the anticipated conditions. This often means selecting a machine that can offer optimum performance in the given conditions, and sometimes adding special features and devices that could provide the flexibility and capability to mitigate adverse ground conditions. In brief, selection of a proper machine with the right technical specifications and functionalities is important to ensure the speedy, safe, and successful completion of the tunnel project within the requirements of the contract. However, by its very nature, subsurface investigation and ground characterization can be all but certain. There are multiple levels of uncertainty involved in ground characterization starting from the location of the borings, formations or layers that could be missed in the drilling program, proper logging and characterization of the lithotypes, accurate measurement of the rock properties in the field and laboratory testing, issues of groundwater and their sources, in situ stresses, and so on. This indicates the need for additional tools for assisting the decision making in the design and construction stage. Typical tools at the engineer’s disposal are statistical methods, systematic risk assessment methods and, more recently, artificial intelligence (AI) approaches such as neural networks and fuzzy logic. Analytic hierarchy process (AHP) is one of the most commonly used multi-criteria decision-making (MCDM) methods, which integrates subjective and personal preferences in performing analyses. However, AHP involves human subjectivity, which introduces a vagueness type of uncertainty. Fuzzy logic, resembling human reasoning in its use of approximate information and certainty to generate decisions, is a better approach to convert linguistic variables to fuzzy numbers under ambiguous assessments, especially in geosciences which suffer from insufficient and uncertain data (KhademiHamidi et al. 2009a). Hence in this study, the vagueness type of uncertainty in the geological information obtained from the site investigation is handled using fuzzy-based techniques to help match the TBM type to the ground conditions. The traditional AHP is modified to fuzzy AHP using fuzzy arithmetic operations, which provides for more flexibility in the application, especially relative to its use in this situation. This paper discusses the methodology and efficacy of the proposed F-AHP in dealing with the selection of the most appropriate rock TBM in adverse geology.
Fuzzy analytic hierarchy process (F-AHP) A fuzzy number describes the relationship between an uncertain quantity x and a membership function lx, which ranges between zero and one. A fuzzy set is an extension of the classical set theory (in which x is either a member of set
123
J. Khademi Hamidi et al.
Coonstrruct the t hieerarcchic tree t
Createe fuzzzy paairwise c parison matrix comp m x
Adjust values
Chheck for consi c istency (C CI) for thee mosst likeely vallue
C CI < 0.1??
Yes C ulatee the fuzzy Calcu f y weiight
Aggreegatee indiividuual A p prefer rencees
Fuuzzy deffuzziffication
Finaal ran nkingg andd deccision n makking Fig. 1 Proposed methodology for fuzzy AHP
A or not) in which an x can be a member of set A with a certain membership function lx. Different shapes of fuzzy numbers are possible (e.g., bell, triangular, trapezoidal, Gaussian, etc.). In order to simplify the implementation of this concept, in this study, triangular fuzzy numbers (TFNs) are used. This paper proposes a seven-step procedure for F-AHP which is schematically given in Fig. 1. Constructing the hierarchical model includes the decomposition of a complex decision problem into smaller manageable elements of different hierarchical levels. The first level of the hierarchy corresponds to objective or goal, and the last level corresponds to the evaluation alternatives (options), whereas the intermediate levels correspond to criteria and sub-criteria. Within a hierarchical structure, the elements of a particular level are compared pairwise with a specific element of an upper level. A fuzzy judgment matrix Je is generated using fuzzy pairwise comparison index e j ij : For n number of comparison items, the fuzzy judgment matrix Je is:
Selection of TMB using fuzzy-AHP
2
j~11
6 6~ eJ ¼ 6 j21 6 . 4 .. j~n1
j~12 .. . .. . j~n2
j~1n
525
3
7 j~2n 7 7 .. 7 . 5 ~ jnn
ð1Þ
It is important to ensure that there is consistency in the pairwise comparisons. Therefore, it would be useful to have a measure of inconsistency associated with the pairwise comparison matrix J. In order to measure the degree of consistency, one can calculate the consistency index (CI). As the next step fuzzy weights are calculated. In this study, for the ease of implementation the geometric mean is adopted to estimate the weights among all various fuzzy weighting techniques. Fuzzy arithmetic operations are utilized over matrix Je to compute the fuzzy weights. The geometric mean is computed for each row Jei : Given Je from Eq. 1, the corresponding fuzzy weights are computed as: 1 Jei ¼ j~i1 j~in n ð2Þ 1 ~i ¼ Jei Je1 Jen ð3Þ w ~i is the fuzzy weight (where i = 1 to n). where w The local priorities at each level are aggregated to obtain final preferences of the alternative. This computation is carried out from the evaluation alternatives to the top level (Goal). Therefore, at each level k of the hierarchical tree, ek are computed as: the fuzzy global preference weight G ek1 ~k G ð4Þ G~k ¼ w The final fuzzy AHP score FeAi for each alternative Ai is obtained by carrying out fuzzy arithmetic sums over each global preference weight: n X ek for each alternative Ai : FeAi ¼ G ð5Þ k¼1
The final fuzzy score for each alternative is then converted into a crisp value through defuzzification to compare and rank them. In this paper the most common centroid index method developed by Yager (1980) is employed. The index is a geometric center xO(Ai) of the fuzzy number of alternative Ai, where for a given TFN (a1, b1, c1) is formulated as follows: R1 Ai lAi ðxÞdx 0 xO ðAi Þ¼ 1 R lAi ðxÞdx 0 ðb1 a1 Þ a1 þ 23ðb1 a1 Þ þ ðc1 b1 Þ b1 þ 13ðc1 b1 Þ ¼ ðb1 a1 Þþ ðc1 b1 Þ ð6Þ where Ai is treated as a moment arm (weight function) measuring the importance of the value x. The value of
xO(Ai) may be seen as the weighted mean value of the fuzzy number Ai. Hence, the bigger the xO(Ai) values are, the better will be the ranking of an alternative.
Application of F-AHP to hard rock TBM selection in adverse geological conditions Rock TBMs There are different classification systems for tunnel boring machines (TBMs), based on different applications, sizes, ground conditions and required final lining. However, depending on the type of the ground excavated, in particular the stand-up time, TBMs can be divided into two main groups; hard rock or soft ground TBMs. Rock TBMs are divided into open TBMs for stable ground and single and double shield TBMs for broken and jointed material. Each type of machine has its own advantages and disadvantages relative to its operation and suitability to cope with anticipated or unanticipated ground conditions. A comparison of three types of rock TBM is given in Table 1. Rock TBM selection in adverse geological conditions Adverse or difficult geological conditions typically refer to situations where the selected TBM cannot work in the operational modes for which it was designed and manufactured. For this reason the TBM advance rate significantly decreases or tunneling may even be stopped. Selection of an appropriate TBM when such adverse and variable conditions are anticipated is very difficult due to potential changes along the tunnel alignment. Therefore, the use of a risk assessment based decision analysis could be a great help in decision making regarding the choice and specifications of the tunneling machine. There are many potential sources of geotechnical problems in mechanized tunneling which should be taken into account when deciding on the selection of a TBM for use in adverse conditions. Among the most significant parameters one can consider is the instability of the tunnel side walls/face, fault zones, squeezing ground, karstic voids, high water inflow and toxic/explosive gases. The main geotechnical hazards relevant to TBM tunneling are given in Table 2, together with the mitigation measures commonly used. These measures can be categorized as precautionary or remedial solutions. However, there have been many TBM tunneling case histories where stopping and changing the tunneling technique was the only possible solution, see for example Barton (2000), Shahriar et al. (2008, 2009). TBMs are mostly site-specific due to project requirements and ground conditions. Parameters to be considered
123
526
J. Khademi Hamidi et al.
Table 1 Comparison of three rock TBMs (after Barla and Pelizza 2000) Double shield TBM
Single shield TBM
Open TBM
Advantages Wide range of application
Wide range of application
Easy operation
Safety
Safety
Applicability in hard rock
Support system flexibility
Precast segmental lining installation
High excavation rate
Simultaneous installation of final support system
High performance
Support system flexibility
Working in falling ground
Working in falling ground
Less construction cost
Controlling water inflow with closed shield
Low investment cost
Disadvantages High investment cost
Two work phases
Grippers inability in unstable rock mass
Complex operation
Drive in weak ground
Support installation in weak rock masses
Need of cleaning the telescopic joint Possibility of TBM jamming in highly convergent ground
Need of precast lining High investment cost Complex operation Need of segment plant
include availability of the manufacturers and timing of the project, used machines available to the contractor, local regulations and standards etc., which may lead to different selection criteria. To demonstrate the fuzzy-AHP approach to rock TBM selection, the Zagros water transfer tunnel was chosen as a case study. The 48 km Zagros tunnel, known as the ‘Nosoud tunnel’, is the longest water transfer project in Iran, extending to a maximum depth of 1000 m although the average depth is 400 m. The project consists of two conveyance tunnels joining along the designated alignment to transfer about 70 m3/s of water in western Iran. The owner is the Iran Water & Power Resources Development Co. (IWPCO). This paper discusses lot #2 of the project which has a length of 26 km and is being constructed using a 6.73 m diameter double shield (DS) TBM. Based on the results of geological site investigations, the main lithological units through which the tunnel will be driven consist of shale, limestone and marl layers (Fig. 2), although the tunnel intersects a variety of geological formations, including Pabdeh, Gurpi and Ilam. The oldest geological unit along the tunnel alignment is the brownish grey limestone of the Ilam formation. The Gurpi formation consists of a combination of limey shale and argillaceous limestone. Some layers in the transition zone between the Ilam and Gurpi formations are rich in pyrites. The youngest unit is the Pabdeh formation which consists of combinations of a dark grey limey shale and greenish grey argillaceous limestone. Frequently changing lithological units and geologic conditions made this project one of the world’s unique and challenging projects. During the driving of the first 9 km of the tunnel, the machine encountered many adverse geological conditions (KhademiHamidi et al. 2008, 2009b; Shahriar et al. 2008, 2009).
123
Shahriar et al. (2008) proposed a methodology to guide decision-making in the selection and evaluation of three rock TBM types for the Zagros tunnel using a geotechnical risk assessment. This approach is now extended to include fuzzy AHP methodology. The geotechnical hazards and their associated risk types, the criteria and sub-criteria of decision-making for rock TBM selection in adverse geological conditions, are given in Table 3. The open, single shield and double shield TBMs are compared for seven major hazards (and their associated risks) including water inflow (WAT), tunnel wall instability (WI), tunnel face instability (FI), fault zone (FZ), karstic voids (KV), squeezing (SQ) and gas seepage (GAS). Each given hazard involves risks related to injuries to personnel, equipment machinery failures and environmental impacts. The risk types are ranked on a scale from zero to three, where a higher value represents a greater risk potential. These risk types are used for pairwise comparison using the proposed F-AHP methodology. The risk assessment results are the basis for the weighting scheme in pairwise comparisons. Based on Table 3, a hierarchical tree of the risks involved in the selection of three different types of rock TBM is illustrated in Fig. 3. In this brief paper, only the pairwise comparison for level 2 criteria (C1–C7) to select the most appropriate TBM type are given in Table 4; the computed fuzzy weights are summarized in Table 5. For a fuzzification factor d = 1, the evaluation of the e3 ¼ w e2 for the three ~3 G final global preference weights, G rock TBMs are summarized in Table 6. The final fuzzy AHP scores FeAi (Table 6) for each alternative (Open, Single shield and Double shield TBM) were evaluated as (0.043, 0.142, 0.486), (0.132, 0.429, 1.391) and (0.132, 0.429, 1.391), respectively. The sum of the most likely values is
Selection of TMB using fuzzy-AHP Table 2 Main geotechnical hazards and their common mitigation measures for TBM tunneling (KhademiHamidi et al. 2009b)
527
Difficult ground conditions
Mitigation measures
Water inflow
Probe drilling Drainage using open or closed channels, as well as pumping the water through drainage pipelines Pre-grouting or grout injection ahead of the face Ground freezing Use of shielded TBMs with bulkhead or pressurized face Use of segmented lining and grouting behind the segments for water tight sealing
Wall instability (often encountered by open TBM)
Use of support systems such as steel arches installed behind cutter head, shotcrete, rock bolts, steel straps, and wire mesh Pretreatment by injection holes Tunnel lining with precast concrete segments Use of shielded TBMs
Face instability
Using fiberglass rock bolt (blocky ground) Creating artificial face Using grill bars in cutter head Consolidation grouting
Fault zones
Probe drilling Ground improvement Segmental lining Ground freezing Drilling drainage holes (high water pressure present) Use of shielded TBMs
Karstic voids
Drilling drainage holes Filling the karstic voids
Squeezing
Over excavation or additional circumferential cutting Use of lubricants such as bentonite, grease Prevention of machine break downs Re-scheduling machine maintenance to avoid long stoppage Use of auxiliary thrust system
Gassy ground
Use of gas monitoring devices and detector systems for the anticipated gas type (i.e., toxic, flammable etc.) Drilling holes in the area of the potential source for gas drainage Use of sufficient ventilation systems to provide fresh air to the TBM and face area Use of gas mask Neutralizing chemical compounds or absorbent material
equal to one (0.142 ? 0.429 ? 0.429), whereas the sum of the minimum values (0.043 ? 0.132 ? 0.132) \ 1 and sum of maximum values (0.486 ? 1.391 ? 1.391) [ 1. The difference between the sum of the minimum values and the sum of the maximum values represents the overall uncertainty (vagueness) in the decision-making process. In the final ranking of the fuzzy AHP score FeAi ; the option with the highest score is ranked the best. Here, the Yager’scentroid index (Yager 1980) is used for defuzzification to rank the alternatives. The final defuzzified values for the Open, Single shield and Double shield TBM are summarized in Table 7. It can be seen that both the Single and Double shield TBMs have
the same rank (preference) and are more desirable options than an Open TBM from the viewpoint of geotechnical risk assessment. This ranking is based solely on the impact of adverse conditions and geological parameters and it shows equal ranking between the two machines based on the weights assigned for the given criteria. If additional criteria were to be added to the set, e.g. speed of tunneling, capability to handle variable ground, cost of machine and operational costs, contractor skills, etc., the outcome of the selection could be different and the ranking of the shielded machines could vary. With the current set of criteria (ground conditions only), the final choice of whether to choose a Single or Double shield TBM depends to a large
123
528
J. Khademi Hamidi et al.
Fig. 2 Longitudinal geological profile of Zagros long tunnel (lot 2)
Table 3 Comparative risk assessment results for rock TBMs
Hazards Water inflow (WAT)
Tunnel walls instability (WI)
Tunnel face instability (FI)
Fault zone (FZ)
Karstic voids (KV) SS TBM single shield TBM, DS TBM double shield TBM 0 no risk because this activity is not involved or negligible value of risk is expected 1 low risk
Squeezing (SQ)
Gas seepage (GAS)
2 medium risk 3 high risk
extent on the contractor’s preference and experience. However, in general, a Double shield TBM is more likely to be chosen, because of its shorter boring cycle and
123
Risk type
Open TBM
SS TBM
DS TBM
Personnel
2
0
0
Equipment
3
1
1
Environment
2
1
1
Personnel
3
0
0
Equipment
3
1
1
Environment
1
0
0
Personnel
1
0
0
Equipment
2
1
1
Environment
1
0
0
Personnel
3
0
0
Equipment
3
2
2
Environment
2
1
1
Personnel
3
0
0
Equipment
3
1
1
Environment Personnel
2 0
1 0
1 0
Equipment
0
1
1
Environment
0
0
0
Personnel
3
2
2
Equipment
2
1
1
Environment
2
2
2
continuous operation which is beneficial to a project of this nature and for a long tunnel. In addition, a Double shield machine could be preferred because of its flexibility when
Selection of TMB using fuzzy-AHP
529
Summary of tunneling operation and selected TBM performance
(C1)
WAT
Personnel (C11) Equipment (C12) Environment (C13)
WI
(C2)
Personnel (C21) Equipment (C22)
Open TBM (A1)
Selecting the most proper TBM in difficult ground conditions
Environment (C23)
FI
(C3)
Personnel (C31) Equipment (C32) Environment (C33)
S. S. TBM (A2)
FZ
(C4)
Personnel (C41) Equipment (C42) Environment (C43)
Equipment (C52)
D. S. TBM (A3)
KV
(C5)
Personnel (C51)
Environment (C53)
SQ
(C6)
Personnel (C61) Equipment (C62) Environment (C63)
(C7)
GAS
Personnel (C71) Equipment (C72) Environment (C73)
Fig. 3 Hierarchical structure for comparison of rock TBMs Table 4 A weighting scheme for major hazards (criteria) C1
C2
C3
C4
C5
C6
C7
~i w
1 2 1=2 3
1=2 1 1=2 2
2 2 1 3
1= 3
1= 2 2 1= 2 3
1= 2 1 1= 2 2
1 2
(0.05, 0.09, 0.20)
1= 2 1=3 1
1 3
1=2 1
2 2
1= 3 1=2
1 2
1= 2 1
2 3
(0.05, 0.12, 0.27)
C6
2 2
C7
1
1=2
1
1= 3
1= 2
1= 3
1
(0.04, 0.08, 0.15)
C1 C2 C3 C4 C5
(0.08, 0.17, 0.35) (0.04, 0.07, 0.17) (0.13, 0.29, 0.57) (0.08, 0.18, 0.37)
encountering difficult ground conditions and the lower degree of risk in tunneling operations associated with these machines.
The Zagros tunnel passes through several sedimentary formations with a wide range of rock mass qualities. During the tunneling operation, changes in rock quality were frequent, with rock masses ranging from poor to very good. The geological conditions encountered changed frequently from hard to soft rock, dry to wet conditions, sticky to non-sticky ground (and vice versa), more often than anticipated by the results of site investigation in the design phase. In the course of the tunneling, the machine has encountered nearly all the adverse geological conditions listed in classical TBM tunneling. The most important problems have been sudden high volumes of water inflowing into the tunnel, noxious hydrogen sulfide and methane gas seepage, and sticky ground, all of which resulted in reduced TBM advance rates. The first incident of a large water inflow was at the beginning of the Ilam limey/carbonaceous formation where a maximum flow of 320 l/s was recorded at the tunnel portal around chainage 4 ? 428 km. Having passed the Ilam formation, the TBM reached the limey and shaly Pabdeh and Gurpi formations where it had been predicted that water inflow would be decreased, due to the lower permeability estimated from the given hydrogeological studies. However, everything was unexpectedly changed when the tunnel reached unit K4Gu in the Gurpi formation at chainage 7 ? 600 km. While the TBM was going through this unit, the maximum water flow at the tunnel portal was 800 l/s, which was exceptional and probably one of the highest inflows recorded to date. According to the engineering geological information, the tunnel face in this unit passes through the syncline S.5 where water pathways (tension fractures and joints) have created a local pressurized aquifer with high permeability. The substantial water inflow into the tunnel caused a number of problems such as face instability, disturbance to segment erection and backfill grouting, material inrush into the front and telescopic shield, pea gravel inrush into the TBM from the annular space behind the lining, disturbance to track-laying ahead of the back-up system, etc., which all in all has led to low TBM utilization. Groundwater may find paths through the tunnel face or walls to enter into the tunnel. With a tunnel excavated by an open TBM, water leakage through the tunnel roof can endanger personnel and electrical installations exposed inside the machine. With a shielded TBM, the shield provides some protection from water flow through the tunnel roof, although some water may leak through telescopic joints or gripper openings. Water movement towards the tunnel also had environmental effects with the groundwater table falling and rate of discharge from springs around the tunnel alignment being reduced. In the Zagros tunnel, most
123
530
J. Khademi Hamidi et al.
~i ði ¼ 1; 2; 3Þ Table 5 Fuzzy local weights for hazards, risk type, w L. 2
W1
C1
(0.05, 0.09, 0.20)
C2
(0.08, 0.17, 0.35)
C3
(0.04, 0.07, 0.17)
C4
(0.13, 0.29, 0.57)
C5
(0.05, 0.12, 0.27)
C6
(0.08, 0.18, 0.37)
C7
(0.04, 0.08, 0.15)
L. 3
W2
W3 (Open TBM)
W3 (SS TBM)
W3 (DS TBM)
C11
(0.49, 0.67, 0.92)
(0.06, 0.07, 0.08)
(042, 0.47, 0.51)
(0.42, 0.47, 0.51)
C12
(0.16, 0.23, 0.33)
(0.06, 0.08, 0.09)
(0.41, 0.46, 0.52)
(0.41, 0.46, 0.52)
C13
(0.07, 0.10, 0.15)
(0.09, 0.11, 0.15)
(0.37, 0.44, 0.53)
(0.37, 0.44, 0.53)
C21
(0.49, 0.67, 0.92)
(0.05, 0.06, 0.07)
(0.43, 0.47, 0.51)
(0.43, 0.47, 0.51)
C22
(0.16, 0.23, 0.33)
(0.06, 0.08, 0.09)
(0.41, 0.46, 0.52)
(0.41, 0.46, 0.52)
C23
(0.07, 0.10, 0.15)
(0.10, 0.14, 0.22)
(0.33, 0.43, 0.54)
(0.33, 0.43, 0.54)
C31
(0.49, 0.67, 0.92)
(0.10, 0.14, 0.22)
(0.33, 0.43, 0.54)
(0.33, 0.43, 0.54)
C32
(0.16, 0.23, 0.33)
(0.09, 0.11, 0.15)
(0.37, 0.44, 0.53)
(0.37, 0.44, 0.53)
C33
(0.07, 0.10, 0.15)
(0.10, 0.14, 0.22)
(0.33, 0.43, 0.54)
(0.33, 0.43, 0.54)
C41
(0.49, 0.67, 0.92)
(0.05, 0.06, 0.07)
(0.43, 0.47, 0.51)
(0.43, 0.47, 0.51)
C42 C43
(0.16, 0.23, 0.33) (0.07, 0.10, 0.15)
(0.07, 0.09, 0.11) (0.09, 0.11, 0.15)
(0.39, 0.45, 0.52) (0.37, 0.44, 0.53)
(0.39, 0.45, 0.52) (0.37, 0.44, 0.53)
C51
(0.49, 0.67, 0.92)
(0.05, 0.06, 0.07)
(0.43, 0.47, 0.51)
(0.43, 0.47, 0.51)
C52
(0.16, 0.23, 0.33)
(0.06, 0.08, 0.09)
(0.41, 0.46, 0.52)
(0.41, 0.46, 0.52)
C53
(0.07, 0.10, 0.15)
(0.09, 0.11, 0.15)
(0.37, 0.44, 0.53)
(0.37, 0.44, 0.53)
C61
(0.49, 0.67, 0.92)
(0.33, 0.33, 0.33)
(0.33, 0.33, 0.33)
(0.33, 0.33, 0.33)
C62
(0.16, 0.23, 0.33)
(0.55, 0.72, 0.91)
(0.12, 0.14, 0.17)
(0.12, 0.14, 0.17)
C63
(0.07, 0.10, 0.15)
(0.33, 0.33, 0.33)
(0.33, 0.33, 0.33)
(0.33, 0.33, 0.33)
C71
(0.49, 0.67, 0.92)
(0.07, 0.09, 0.11)
(0.39, 0.45, 0.52)
(0.39, 0.45, 0.52)
C72
(0.16, 0.23, 0.33)
(0.09, 0.11, 0.15)
(0.37, 0.44, 0.53)
(0.37, 0.44, 0.53)
C73
(0.07, 0.10, 0.15)
(0.33, 0.33, 0.33)
(0.33, 0.33, 0.33)
(0.33, 0.33, 0.33)
e3 ¼ w e2 ~3 G Table 6 Evaluation of final global preference weights for the rock TBMs, G Hazards
Risk type
Open TBM
SS TBM
DS TBM
Water inflow
Personnel
(0.001, 0.004, 0.014)
(0.009, 0.028, 0.094)
(0.009, 0.028, 0.094)
Equipment
(0.000, 0.002, 0.006)
(0.003, 0.009, 0.034)
(0.003, 0.009, 0.034)
Environment
(0.000, 0.001, 0.004)
(0.001, 0.004, 0.016)
(0.001, 0.004, 0.016)
Personnel
(0.002, 0.007, 0.022)
(0.016, 0.054, 0.166)
(0.016, 0.054, 0.166)
Equipment
(0.001, 0.003, 0.011)
(0.005, 0.018, 0.059)
(0.005, 0.018, 0.059)
Environment
(0.001, 0.002, 0.011)
(0.002, 0.007, 0.028)
(0.002, 0.007, 0.028)
Personnel
(0.002, 0.007, 0.034)
(0.006, 0.021, 0.086)
(0.006, 0.021, 0.086)
Equipment
(0.001, 0.002, 0.008)
(0.002, 0.007, 0.030)
(0.002, 0.007, 0.030)
Environment Personnel
(0.000, 0.001, 0.005) (0.003, 0.011, 0.035)
(0.001, 0.003, 0.014) (0.028, 0.092, 0.268)
(0.001, 0.003, 0.014) (0.028, 0.092, 0.268)
Equipment
(0.002, 0.006, 0.021)
(0.008, 0.030, 0.097)
(0.008, 0.030, 0.097)
Environment
(0.001, 0.003, 0.013)
(0.004, 0.013, 0.045)
(0.004, 0.013, 0.045)
Personnel
(0.001, 0.005, 0.017)
(0.011, 0.038, 0.129)
(0.011, 0.038, 0.129)
Equipment
(0.001, 0.002, 0.008)
(0.003, 0.012, 0.046)
(0.003, 0.012, 0.046)
Environment
(0.000, 0.001, 0.006)
(0.001, 0.005, 0.021)
(0.001, 0.005, 0.021)
Personnel
(0.014, 0.041, 0.112)
(0.014, 0.041, 0.112)
(0.014, 0.041, 0.112)
Equipment
(0.007, 0.029, 0.109)
(0.002, 0.006, 0.021)
(0.002, 0.006, 0.021)
Environment
(0.002, 0.006, 0.018)
(0.002, 0.006, 0.018)
(0.002, 0.006, 0.018)
Personnel
(0.002, 0.005, 0.016)
(0.008, 0.023, 0.074)
(0.008, 0.023, 0.074)
Equipment
(0.001, 0.002, 0.007)
(0.003, 0.008, 0.027)
(0.003, 0.008, 0.027)
Environment
(0.001, 0.003, 0.008)
(0.001, 0.003, 0.008)
(0.001, 0.003, 0.008)
(0.043, 0.142, 0.486)
(0.132, 0.429, 1.391)
(0.132, 0.429, 1.391)
Tunnel walls instability
Tunnel face instability
Fault zone
Karstic voids
Squeezing
Gas seepage
FeAi
P
e3 G
123
Selection of TMB using fuzzy-AHP
531
Table 7 Ranking of rock TBMs using defuzzification method
Table 8 The best records of TBM performance parameters during the project life (Khademi Hamidi 2009)
Alternative
Centroid, xo(Ai)
Rank
Open TBM
0.224
2
Maximum weekly advance rate (m)
190.53
Single shield TBM
0.651
1
Maximum monthly advance rate (m)
757.95
Double shield TBM
0.651
1
Maximum rate of penetration (m/h)
Maximum daily advance rate (m)
Maximum daily utilization factor (%)
of the springs dried up completely, even at a distance of more than 4 km from tunnel axis. In addition to high groundwater inflows, there were high gas emissions from the Ilam formation which caused serious problems for the tunnel crew and TBM components. In extreme conditions more than 100 ppm H2S was recorded, accompanied by a huge amount of groundwater flowing into tunnel and resulting in a four-month shutdown of tunneling operations. Hydrogen sulfide is highly soluble in water, resulting in a moderately acidic solution which seeped into the tunnel causing problems with corrosion of the TBM components as well as environmental impacts when it was released into the atmosphere. In addition, on several occasions the gas detection and measuring sensors in different parts of TBM shield and back-up system recorded concentrations of CH4 up to 100% of the lower explosive limit (LEL), resulting in several tunnel evacuations and one gas explosion accident in the cutter head area. The presence of such a high gas concentration in the tunnel was related to the iron sulfide minerals often found in such geological formations as the Pabdeh and Gurpi, which are the major natural oil (gas)-bearing basins of western Iran. To cope with such adverse and complex geological conditions and their impact on the tunneling operation, different measures were recommended and taken into account with the continuous collaboration of the contractor and consultants (KhademiHamidi et al. 2009b; Shahriar et al. 2009). As a consequence of the mitigation plans provided for a safer work environment and improved working conditions, the tunneling proceeded at higher utilization rates and an overall higher daily advance rate. Monthly advance rate (m)
9000
Cumulative advance rate (m)
8000
Monthly advance (m)
700 600
High water inflow and gas emission accident
500
TBM shutdown
Second high 7000 water inflow 6000 5000
400 4000 300 TBM restart 200
3000 2000
100
Cumulative advance (m)
800
34.78
3.63 57
Table 9 The TBM performance parameters during the year 2006 (Khademi Hamidi 2009) Maximum daily advance rate (m) Average daily advance rate (m) Maximum weekly advance rate (m) Average weekly advance rate (m)
30.55 10.30 177.08 45.86
Maximum monthly advance rate (m)
591.41
Average monthly advance rate (m)
354.83
Maximum rate of penetration (m/h)
3.63
Average rate of penetration (m/h)
2.75
Maximum utilization factor (%)
44
Average utilization for the period (%)
15.6
Figure 4 indicates the available monthly and cumulative advance rates of the TBM. As can be seen, except for the months in which the TBM encountered gas and groundwater problems, the TBM advance rate varied quite widely but overall was acceptable and consistent with the norms for TBM tunneling with a similar size machine. Tables 8 and 9 provide a summary of the TBM performance parameters during the project life and the year 2006, respectively. The work was based on 3 shifts a day, 7 days a week. As can be seen from Table 9, the TBM performance parameters are satisfactory to January 2007, i.e. before any serious accident. The overall performance of the machine illustrates that despite some of the unforeseen difficulties and interruptions, the selected TBM is an appropriate choice for this project. In the light of the experience gained from the Zagros tunnel, for future projects a refined method of assigning various types of risks and weighing system can be used in the F-AHP method to assist engineers in the process of machine selection. It could also include additional performance and operational criteria that would allow more quantitative differentiation between various choices of machines and more specifically their trailing gears and additional accessories.
1000
0 Mar-06 Apr-06 May-06 Jun-06 Jul-06 Aug-06 Sep-06 Oct-06 Nov-06 Dec-06 Jan-07 Feb-07 Mar-07 Apr-07 May-07 Jun-07 Jul-07 Aug-07 Sep-07 Oct-07 Nov-07 Dec-07 Jan-08 Feb-08 Mar-08 Apr-08 May-08 Jun-08 Jul-08 Aug-08 Sep-08 Oct-08 Nov-08 Dec-08 Jan-09 Feb-09
0
Fig. 4 TBM monthly and cumulative advance rate during Zagros tunnel life
Conclusions In view of the variability of the ground conditions through which the Zagros tunnel was to be driven, a fuzzy-AHP
123
532
methodology was used to study the selection of the most appropriate tunneling machine. The F-AHP used fuzzy arithmetic operations to aggregate the fuzzy global preference weights with respect to each alternative. The criteria chosen were water inflow, tunnel wall and face instability, fault zone, karstic voids, squeezing and gas seepage, which are the most common hazards encountered when a TBM is advancing through adverse geological conditions. The alternatives were ranked based on the score estimated by the summation of final global preference weights. The final defuzzified scores for an Open, Single shield, and Double shield TBM indicate that under the given assumptions the shielded TBM (either Single or Double) was preferable to the Open TBM. However, a Double shield TBM has more flexibility in difficult ground conditions and is likely to provide higher advance rates, which is particularly important when driving longer tunnels. The advance rates recorded for the Zagros tunneling machine during excavation of the completed sections validate the TBM selection process and confirm that the FAHP method is a useful tool to aid decision-making in the selection of a TBM to drive a tunnel through a heterogeneous/complex rock mass. In the current study the choices were between three different machine types such that experienced tunnel engineers and machine experts could analyze the information and recommend a machine type. However, F-AHP method can be used to its full capacity when it is applied to the selection of additional components and accessories for the tunneling machine where there are multiple levels of decision-making parameters and
123
J. Khademi Hamidi et al.
consequences that cannot be easily accounted for unless a systematic method is utilized.
References Barla G, Pelizza S (2000) TBM tunnelling in difficult ground conditions. Geoeng, Melbourne Barton N (2000) TBM tunneling in jointed and faulted rock. Balkema, Brookfield, p 173 KhademiHamidi J (2009) A model for hard rock TBM performance prediction. Dissertation, Amirkabir University of Technology (in preparation) KhademiHamidi J, Bejari H, Shahriar K (2008) Assessment of ground squeezing and ground pressure imposed on TBM shield. Proceedings 12th international conference of international association for computer methods and advances in geomechanics, Goa, India KhademiHamidi J, Shahriar K, Rezai B, Bejari H (2009a) Application of fuzzy set theory to rock engineering classification systems: an illustration of the rock mass excavability index. Rock Mech Rock Eng. doi:10.1007/s00603-009-0029-1 KhademiHamidi J, Shahriar K, Rostami J (2009b) Double shield TBM in challenging difficult ground conditions: a case study from Zagros long water transfer tunnel, Iran. Proceedings of the rapid excavation and tunneling conference (RETC). Las Vegas, Nevada, pp 1321–1333 Shahriar K, Sharifzadeh M, KhademiHamidi J (2008) Geotechnical risk assessment based approach for rock TBM selection in difficult ground conditions. Tunnell Undergr Space Technol 23:318–325 Shahriar K, Rostami J, KhademiHamidi J (2009) TBM tunneling and analysis of high gas emission accident in Zagros long tunnel. Proceedings of the ITA-AITES World Tunnel Congress 2009 and 35th ITA general assembly, Budapest, Hungary Yager RR (1980) On a general class of fuzzy connectives. Fuzzy Set Syst 4:235–242
